Technical insights on AI, quantitative finance, and trading systems
Turning a mobile end agent from a demo into a product users can trust and pay for requires engineering the fundamentals: scalable concurrency, low-latency execution, explicit safety guardrails, and sustainable token-level unit economics.
Notes from building a low-latency voice-to-video pipeline that turns spoken stories into projected cinematic sequences: streaming ASR, LLM scene planning, keyframes, motion generation, and a projector-ready render loop.
Building a RAG agent for quantamental research using Voyage AI embeddings, Qdrant vector store, Exa web search, and DeepSeek reasoning. Combines proprietary research with real-time market intelligence.
Exploring SFT and RL methods to train models on options pricing, Greeks calculation, volatility surface modeling, and hedging optimization. Integrated with VaR, stress testing, and Tinker API for complex derivative scenarios.
An introduction to building AI agent systems for quantitative trading. This article explores agent architectures, LLM integration patterns, and evaluation frameworks for financial applications.